We have developed a novel method for producing chimeric mice with fully mature and functional human liver tissue, which is maintained in the absence of any ongoing liver pathology and without continued drug treatment. This model system overcomes the substantial problems (kidney and liver pathology, short lifespan, incomplete liver replacement) that have significantly limited the utility of existing models. To do this, a highly immunocompromised (NOG) mouse was genetically engineered to express a thymidine kinase transgene in the liver, which enabled the temporally controlled and tissue-specific ablation of mouse liver tissue after administration of a non-toxic drug (gancyclovir). After injection of human liver cells, the recipient mice were shown to have a high level of human liver reconstitution in situ, which survived for a more prolonged period (>8 months) than could be attained in any other model;and to have a gene expression profile and 3-dimensional architecture characteristic of mature human liver. First, this platform will be used to develop novel transplantation methods that enable human liver stem cells, human induced pluripotent epithelial cells (hiPS), or human adipocyte-derived stem cells to reconstitute a human liver. If successful, this would enable a complete paradigm-shift for treatment of end-stage liver disease: 'liver transplantation'could be performed using autologous cells without immunosuppression. Secondly, by genotyping the donor human liver cells, a panel of mice with 'humanized livers'with specified alleles for phase I drug metabolizing will be produced. Then, the rate of formation of human-specific drug metabolites for at least 3 drugs, whose metabolism is dependent upon human CYP450 enzyme activity, will be analyzed. If successful, this innovative model system will enable pharmacogenetic analyses to be efficiently performed within an entirely human context in vivo with all confounding environmental variables controlled, which would overcome a significant roadblock for characterizing inter-individual variation in drug metabolism.